Chris M Evans

Chris M Evans

@chrismevans.me

Owner, founder and Principal Analyst at Architecting IT. Food lover, eater and maker. Husband, father. Lover of all things computing (except perhaps Networking). https://architecting.it

265 followers157 following858 posts

Building data centres sounds a little more challenging (I'm thinking of the wiring/cabling issues), but we build cars with robots, so why not? (4/5)

The use of robots would seem a logical approach to scaling data centre footprints. Robots can move around 24/7 carry many spare parts, travel up and down data centre aisles easily and recharge themselves automatically. But is this a $100 billion business? (5/5)

We already have robot warehouse automation & manufacturing, so data centre build and operation doesn't seem a great leap, although I question the valuation being suggested. (2/5)

Data centres are pretty modular by design, so I can see an initial benefit of using robots to service (remove/replace) failed equipment. With even greater use of modularisation & standardisation, this process should be easy for robots. (3/5)

The PC market is wider than just two players. I don't see Dell as relevant, due to the legacy parts of the business that NVIDIA will not want and need to dispose of. (3/4)

However, I do see NVIDIA acquiring a PC vendor in order to sell its upcoming N1 chips. The market is competitive and NVIDIA needs a good channel to customers for its new processor. So, despite denials, an acquisition could be on the cards, but I don't think it will be a legacy brand. (4/4)

What rings alarm bells in this article is the difference between Dell and HP in terms of market cap and market coverage. Dell is about $122.5 billion with enterprise & consumer divisions. HP is valued at about $17.5 billion and of course, was the PC spinoff from HPE. (2/4)

As a man who has already ended eight wars, the current one shouldn't be hard for Trump to finish - after all, he started it. After massive tariffs on the UK, insulting our soldiers and saying he didn't need our ships, now he asks for help. www.bbc.co.uk/news/live/cx...

bbc.co.ukTrump says he's 'not happy' with UK after Starmer says it won't be drawn into wider Iran warThe US president repeats his call for other nations to assist in securing the Strait of Hormuz, after saying it would be

Fairly sure Pete Hegseth is talking bollocks here. The General Belgrano was sunk by submarine torpedoes during the Falklands War in 1982.

Currently watching the Western Digital Innovation Day livestream (unfortunately joining late). I think the CFO is onstage crowing about the remaining shares in Sandisk they own, which are currently worth about $5 billion.

However, Sandisk is currently worth $96 billion and the company owned 100% of it just before the split. At the split, when Sandisk started trading it was worth about $5.6 billion. WD just gave away $90 billion of value and they seem happy with that.

Orbital data centres? Not sure I see the science or logic behind this, other than the ability to avoid any kind of regulation. So many questions need to be asked here. How long will they be in space, how big will they be? Do they just get destroyed on re-entry? Where's the sustainability? (3/4)

However, despite the negativity, Musk is a genius at keeping the rollercoaster going. Whether that's the dream of orbital data centres or the dream of millions of humanoid robots. He takes Steve Jobs' reality distortion field to another level. (4/4)

It's the subprime mortgage issue all over again, but perhaps not as toxic. Existing X & xAI investors will be happy as they get a slice of a better quality company that has government contracts (SpaceX) and the Starlink business. (2/4)

View full thread

A third "long shot" option is that current sovereignty issues create a rethink on using Windows as a desktop O/S, due to the invasive nature of the current requirements to register an online account. (8/9)

Unfortunately, I think this is likely to be a strategy taken up by only those hard-core tech fans and not the general public. Lots of interesting things to mull over, as usual! (9/9)

View full thread

Ultimately, that's the crux of the issue for Arm-based desktops outside of Apple. Yes, Linux desktops are amazing and there are many to choose from. But (a) they're not as slick as Windows for installation (b) they don't have application support to equal Windows. (6/9)

Two scenarios exist - first, Windows is improved to operate more efficiently on Arm than x86; second, apps support gets good enough on non-Windows operating systems to make Linux viable. (7/9)

View full thread

It sold off the storage division. All to invest in Foundry, which continues to make massive losses ($2.5 billion in the latest quarter alone). (4/9)

Arm is an alternative for the desktop that may see more popularity if the price point addresses issues like the cost of memory and storage. Apple Arm-based devices need less memory, although they do run a proprietary O/S. The remainder of the market is effectively stuck with Windows. (5/9)

Once again Intel risks the Innovator's Dilemma scenario. It overplayed the concept of the "AI PC", which nobody wanted. It stepped away from the storage industry, when products like Optane could have plugged some of the current memory gap. (3/9)